Studies of lymphocyte dynamics in lymphopenic environments such as the neonatal mouse or in cell transfer into genetically lymphopenic (Rag2-/-; CD3e-/-) recipients indicate that a striking proliferative response occurs on the part of some of the transferred cells and that such cells undergo 7 or more cell divisions within one week and display a memory phenotype at the end of this very striking response. In previous work, it was shown that the pre-transfer of CD4 T cells into lymphopenic recipients would prevent newly introduced cells from undergoing rapid proliferation. It was also shown that the capacity of the initially transferred cells to block proliferation by subsequently introduced cells correlated very strongly with the TCR repertoire complexity of the cells present as the time of the second transfer. Thus, although Rag2-/- mice that had received 30,000 or 2 million CD4 T cells had similar numbers of memory phenotype CD4 T cells 6 weeks after transfer, newly introduced cells showed striking proliferation in a host that had previously received 30,000 cells but failed to proliferate in a host that had previously received 2 million cells. The TCR complexity was far greater in the recipients of 2 million cells than in the recipients of 30,000 cells despite the similar numbers of cells present at equilibrium. [unreadable] Unit scientists observed that recipients of small numbers of CD4 T cells eventually developed a fulminant macrophage/ eosinophil pneumonia, characterized by the presence of alternatively activated macrophages, Ym1 crystals and eosinophils. These mice also showed an eosinophilic gastritis and a striking increase in the frequency of cells that produced IL-4, IL-13 and IL-5 in the lymph nodes. This syndrome bears a striking relationship to that observed in humans in which there is a limitation of thymic supply of T cells to the periphery, such as Omenn's syndrome, atypical complete DiGeorge syndrome and maternal engraftment in severe combined immunodeficiency disease. If the transfers are carried out in mice that lack T cells but express B cells (CD3e-/- mice), profound induction of serum IgE is observed. [unreadable] It has now been shown that pre-transfer of regulatory T cells will block the induction of the eosinophilic inflammatory disease induced by transfer of limited numbers of conventional CD4 T cells. However, this is observed only when the number of initially transferred regulatory T cells is relatively large (300,000); pre-transfer of 30,000 regulatory T cells does not block induction of disease even though the number of regulatory T cells at the time of the introduction of the conventional T cells is the same (as a result of homeostatic expansion). This implies that regulatory T cells must have a sufficiently complex TCR repertoire to allow them to block activation of conventional T cells and of disease induction mediated by these cells. Rigorous depletion of Treg cells from transferred populations has been used to determine whether the preference of Th2 differentiation when precursor number is low is simply due to the presence of fewer T regs. However, making use of cells from Foxp3 indicator mice, it seems clear that this is not the case and cell number is a critical deterministic issue. A strain of mice with limited repertoires based on their possessing only a single VbDbJb chain and a single Va have also been shown to develop severe eosinophilic inflammatory disease with involvement of the lung. Thus, limited repertoires achieved in more than one manner have similar implications for the development of autoinflammatory disease.[unreadable] Studies utilizing mice transgenic for the TCRb chain used in the 5C.C7 anti-cytochrome C peptide T cell receptor have allowed a test of the propensity of cells stimulated under strong and weak signaling to preferentially generate Th1 or Th2 responses. Using a variety of tools, including relative expression of CD5 and avidity of binding to cyctochrome C peptide tetramers, it is clear that low signal strength stimulation favors Th2 differentiation in vitro and, in a set of limited tests, in vivo as well.